nexuslinkproductions/yuri-os - 第 21 页

gpt-oss-local-runtime

Set up and reason about local gpt-oss runtime paths, defaulting to Ollama with LM Studio as fallback. Use when you need a concise local-offload plan, endpoint mapping, or compatibility notes for OpenAI-shaped clients.

原文语言：英语

grill-me

Interview the user relentlessly about a plan or design until reaching shared understanding, resolving each branch of the decision tree. Use when user wants to stress-test a plan, get grilled on their design, or mentions "grill me".

原文语言：英语

grill-with-docs

Grilling session that challenges your plan against the existing domain model, sharpens terminology, and updates documentation (CONTEXT.md, ADRs) inline as decisions crystallise. Use when user wants to stress-test a plan against their project's language and…

原文语言：英语

haki-intent

Intent Pre-cognition Engine — builds a probability-ranked map of the 5 most likely underlying intents, surfacing hidden constraints (P4) and meta-needs (P5). Use when a request's true intent is ambiguous or high-stakes, or when the user invokes /haki.

原文语言：英语

hatch-pet

Create, repair, validate, visually QA, and package Codex-compatible animated pets and pet spritesheets from character art, generated images, company or prospect brand cues, or visual references. Use when a user wants a lightweight-worker Codex pet workflow, a…

原文语言：英语

humanizer

Edit user-supplied prose to remove generic AI-writing patterns while preserving meaning, evidence, citations, and the writer's own voice. Use when the user asks to humanize, de-slop, polish, or voice-match prose.

原文语言：英语

imagegen

Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or…

原文语言：英语

improve-codebase-architecture

Find deepening opportunities in a codebase, informed by the domain language in CONTEXT.md and the decisions in docs/adr/. Use when the user wants to improve architecture, find refactoring opportunities, consolidate tightly-coupled modules, or make a codebase…

原文语言：英语

izanagi-simulator

Counterfactual simulation engine that generates 3 divergent alternate paths before committing to high-stakes plans, evaluating each by EV/risk/reversibility with explicit simulation records. Use when facing CRITICAL or HIGH complexity decisions with multiple…

原文语言：英语

kimi-k2-6-server-adapter

Plan Kimi K2.6 as a self-hosted, server-grade model behind an OpenAI-compatible endpoint. Use when you need a concise adapter brief, not a consumer-local runtime plan.

原文语言：英语

math-curve-loaders

Reference a read-only catalog of math-based loading animations, formulas, and motion patterns. Use when you need a catalog skill, not a model-routing or runtime skill.

原文语言：英语

migrate-to-shoehorn

Migrate test files from `as` type assertions to @total-typescript/shoehorn. Use when user mentions shoehorn, wants to replace `as` in tests, or needs partial test data.

原文语言：英语

mineru-document-extractor

Document extraction tool using MinerU CLI to convert PDFs and other documents to structured Markdown. Use when the user says "extract this document", "convert PDF to markdown", "parse this PDF", "I need text from this document", or when needing to extract…

原文语言：英语

mure-advisor

[CONSULT-ONLY; not operational or dispatch authority; Advisory context only until canonical MURE role cards and executable topology are recovered by task t-63a57bfd.] Tiered per-turn advisor for main session + WhatsApp. Cheap watcher + heavy escalation. Pure…

原文语言：英语

mure-role-variant-matrix

[QUARANTINED STALE DOCTRINE; model auto-selection prohibited; Declares retired OpenClaw residue authoritative; preserved for explicit read-only archaeology until MURE task t-63a57bfd repairs it.] MURE role × model variant matrix. One role, several tuned…

原文语言：英语

nen-phase-detector

Adaptive phase specialization that sets the work-phase mode (deep_code, design, strategy, research, admin, rest) to tune depth, verbosity, and approach to match what the current phase needs. Use when starting a new work phase, when the user says "/nen…

原文语言：英语

nex-deliverables

Client deliverable tracking and project management system for web agencies, design studios, marketing firms, and freelancers managing multiple simultaneous projects and client relationships. Use when the user says "track deliverables", "what's overdue", "send…

原文语言：英语

nex-vault

Secure local contract and document vault for managing business agreements with automatic expiration tracking, renewal deadline alerts, and compliance monitoring. Use when the user says "track my contracts", "when does this lease expire", "add a new contract",…

原文语言：英语

nexus-security-hardening

Wave-2 security hardening for Nexus Link — runs the diagnostic tools, dispatches the MURE build, verifies the gates. Invoke for "wave 2", "security hardening", "nexus security", "pre-customer security", "rls audit", "audit chain verify", "antipattern scan".

原文语言：英语

non-destructive-infinity-guard

Always-on action boundary, risk classifier, and mutation approval gate for Yuri OS. Use when a user or agent invokes /yuri guard, /guard, or /ndig, or when any proposed action, tool call, or mutation needs risk classification, scope verification, and approval…

原文语言：英语

objection-source-diagnoser

Diagnose WHY a deal is accumulating objections by tracing each one back to its root-cause seller behavior. Use this skill when a customer keeps pushing back, when you're getting too many price objections, when the prospect raised concerns you don't know how…

原文语言：英语

openai-codex-workflow

Research and apply OpenAI/Codex workflow guidance, including docs, config, subagents, skills, memory, and local gpt-oss usage. Use when you need a compact, reusable brief for how to work safely and effectively with OpenAI agent tools.

原文语言：英语

openai-docs

Use when the user asks how to build with OpenAI products or APIs and needs up-to-date official documentation with citations, help choosing the latest model for a use case, or model upgrade and prompt-upgrade guidance; prioritize OpenAI docs MCP tools, use…

原文语言：英语

opus-fleet

[CONSULT-ONLY; not operational or dispatch authority; Substrate advice only; it is not current MURE role, topology, or provider-binding authority pending task t-63a57bfd.] Standing operating model — Opus (and its glm-5.2 Opus-equivalent) orchestrates while…

原文语言：英语

oracle-adapters

Build compatibility adapters between Jarvis, OVOS, OpenClaw, and Oracle-native surfaces. Use when legacy triggers or old labels must be preserved while the visible OS surface is cleaned up.

原文语言：英语

oracle-memory

Build Oracle memory surfaces: session logs, durable notes, local context, and retrieval. Use when the task touches historical context, research notes, or project memory on this OS.

原文语言：英语

oracle-registry

Build Oracle registry surfaces for skills, packages, and installs. Use when the task touches local skill packs, discovery, publishing, or install/update flow on this OS.

原文语言：英语

oracle-router

Build Oracle task routing: session ownership, detached work, launch order, and handoff rules. Use when the task touches multi-step execution or agent swarm coordination on this OS.

原文语言：英语

oracle-voice

Oracle audio-subsystem builder: STT, TTS, wakeword, call handoff, and audio routing for the Oracle shell. Invoke by name (oracle-voice) only when building the Oracle audio pipeline. NOTE: not the native dictation command — for Claude Code dictation use the…

原文语言：英语

organ-discovery-precision-gate

Gates a lane claim against its WorkSubstrate scope and discovery footprint before the energy gate runs, verifying the lane stayed inside its granted authority. Use when a lane produces a claim that needs scope-and-footprint verification, or when invoking…

原文语言：英语

organ-filing-assessor

Deterministic, read-only placement assessor that classifies YURI artifacts into canonical zones, scores purge candidates by hazard-decay staleness, and deduplicates. Use when generated artifacts (research docs, memories, reports, lane outputs, telemetry) need…

原文语言：英语

organ-formula-foundry-bakeoff

Foundry Core B bakeoff harness and promotion/demotion ledger that runs synthesized formula candidates up a staged promotion ladder and records every gate verdict to an append-only ledger. Use when formula candidates need scoring, promotion, or demotion…

原文语言：英语

organ-formula-foundry

Formula Foundry typing Core A — the legal-move generator that classifies formula input/output units into closed-set dimensions and rejects dimensionally-incoherent compositions. Use when composing formulas and needing to verify dimensional compatibility,…

原文语言：英语

organ-lane-telemetry-cockpit

Human-readable cockpit that reads originator-telemetry.jsonl, groups by traceId, and prints a compact per-run timeline with elapsed time, heartbeat count, chars streamed, and status. Use when you need to see what a lane is doing without reading raw JSONL, or…

原文语言：英语

organ-openprocess-pool

Mathematical memory for started-but-unclosed work that ranks open tasks, research, ideas, todos, skills, bugs, experiments, and handoffs by open mass (status, staleness, dependency centrality, risk, and value). Use when you need a math-grounded answer to…

原文语言：英语

organ-yuri-decode

LLM-wielded decoder instrument: translate text into a deterministic math object (tokens, numerology channels, dimensional classification). Use when the model needs a deterministic structural reading of text to reason over — e.g. decoding a brain dump's…

原文语言：英语

organ-yuri-nerve

Nervous-system spine + afferent nerve: captures detected events once under a deterministic id and threads them across organs as OpenProcess entries. Use when durably recording an event (task/decision/fix/finding/idea/correction/build/handoff) so it survives…

原文语言：英语

parallel-clone-orchestrator

Retired skill tombstone. Do not invoke — use native planning plus llm-compat lanes for advisory work instead. Use when encountering a reference to /pco, /clone, or parallel-clone-orchestrator and need to redirect to the current replacement: native planning +…

原文语言：英语

pattern-mirror-core

Artifact perception, pattern extraction, weakness detection, and YURI-native reconstruction — inspired by Sharingan/Copy Technique, translated into enterprise-safe system behavior. Use when analyzing and integrating an external artifact (repo, document, PDF,…

原文语言：英语

peer-signal-build

Curated recipe for a PEER LANE (native Sonnet agent OR ollama nano-swarm) to build ONE alpha-factor-library signal/scoring/analysis module for the crypto trading engine — exact contract, signal interface, fail-open discipline, --test, capability tag,…

原文语言：英语